-
Notifications
You must be signed in to change notification settings - Fork 2
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Split torch.Tensor and metatensor bindings #4
Conversation
db9725b
to
3349401
Compare
3349401
to
e1f10cc
Compare
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I liked most of the changes. Since meshlode.lib.System
is (for now) one of the user-facing functions, I have now tried to homogenize the UI a little by importing directly from meshlode import System
. Apart from one potential bug that I comment on below, the code looks fine to me.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
When I run this file, I see that the sign of the obtained energies is different for the torch and metatensor versions. This should probably be consistent.
e75bc19
to
3b7ed81
Compare
6422ee3
to
8e454b8
Compare
As we discussed the last time: I split the
MeshPotential
into one class that returns a list of torch Tensors and now lives incalculators
. Also there is another one (with the same name) that lives inmetatensor
. Let me know is you agree on this structure?I also refactored the docs a bit and moved the
FourierSpaceConvolution
andMeshInterpolator
andSystem
class into alib
folder.TODOs